Brahmaputra Floodplains: Inundation Extent Prediction

Briksha Borna Gogoi

GE4007

This project aims to predict flood and it's extent of inundation in the Brahmaputra River's floodplains, focusing on a 140-km stretch in Assam, India. Utilizing HEC-HMS for rainfall-runoff modeling and HEC-RAS for Unsteady Flow Simulations, it maps simulation for different Return Periods and intensity, understanding historical flood patterns (2015-2024) using Sentinel-1 SAR imagery and predicts 2025 inundation with Random Forest ensemble learning. The model estimates a 100-year rainfall intensity of 7.85 mm/h (2025), forecasting extensive flooding affecting 2285.87 km². By comparing SAR and HEC-RAS outputs, the project identifies vulnerable villages, enhancing disaster preparedness and supporting sustainable development in flood-prone regions.

Report Content

Introduction to Flood, its Causes and Need of study

Aim & Objective, Defining Study Area

Methodology and Dataset

Hydrologic Analysis: Generation of Hydrographs to understand peak run-off using HEC-HMS

Hydrologic Analysis: Unsteady Flow Simulation in HEC-RAS for different Return Periods

Deriving historical Flood maps through high resolution SAR images and Flood Frequency Analysis

Validation of HEC-RAS simulation with SAR derived flood areas layer: Comparison for 2024

Prediction Outputs: Random Forest and HEC-RAS

Method Comparison for 2025

Probable Areas that would be affected and Conclusion